Testpage3
Project Title: Battlefront III Developer: Free Radical Design Milestone: Milestone 5 Production: Ted Fitzgerald Jamil Dawsari Gabe Gils Carbo James Morris Milestone Review PC Version October 26ᵗʰ, 2006 Milestone 6 - Monday 23ʳᵈ October 2006 Running on PC and X360: *Jedi Melee (in-game to evaluate gameplay) *Jedi Force Power (Lighting and Choke) prototype only affecting characters * AI is functioning on ground/in space (ability to follow player across VBF) *All classes & playable units fully defined including attributes, weapons and role *Flying AI taking off / and landing on terrain and capital ships **Can fly from ground to capital ship and vice versa * Destruction of vehicles with debris with placeholder effects * FX – re-entry in-game * FX - Lightsaber in-game ms6.png Build Performance Overall *Milestone working on PS3, 360 and PC *Nicely done getting MS6 onto the PS3 Good *Good stability with 2-screen multiplayer *PS3 Jedi Melee running at 60 fps Areas to Improve *'B Issue': Issues with stability and low framerate affecting build – what plans are there to improve framerate by MS7? ** This will be covered by FRD’s optimization plan ** LEC – that’s great, what’s your timeline for delivering the optimization plan? ** FRD: We will discuss this next week, or in the conference call I have requested with Peter and Steve *'C Issue': Client/Server desyncs ** Noted.Thanks for all information and emails. We will look into it. *'B Issue': Unable to load Mos Espa single-player without +perf command parameter (PC only) *'B Issue': Mos Espa single-player @ 3-5fps (all platforms) *'B Issue': Unstable single-player (PC only) **When we execute game.win.exe without commandline options and load up Mos Espa, the game crashes or locks up within a few seconds of either loading the spawn menu, or a few seconds after spawning into the world. (PC Only) ** These last three B issues – Is this applicable to all platforms? The PC version is currently targeted to one specification. Running on hardware that differs from it is likely to produce issues until compatibility testing and fixing is carried out ** LEC – No, as noted the +perf and crash were PC specific. The low frame-rate was across all platforms on Mos Espa. ** FRD: We will investigate. Please can I have the spec of the worse machines. STATUS: Acceptable given milestone, will not impact approval Jedi Melee (in-game to evaluate gameplay) Overall *Good first pass on Jedi Melee Good *Locking lightsabers *Blocking animations *Attack animations Areas to Improve *'B Issue': Target Lock-on: Players need a mechanism for turning off the lock. ** Targeting lock occurs when the enemy is within a certain range. Unlock occurs if the enemy goes outside the range. Presently this works as designed ** LEC – So, if the player is moving through a crowd of AI trying to reach the enemy hero he could find himself constantly locking onto the AI and being frustrated. The BF2 system of locking on automatically when a target is attacked, but also having the ability to turn lock on/off allowed for more user freedom and less frustration. ** FRD: To clarify, the lock-on will only happen if you choose to attack. Let’s discuss further on your visit. *'B Issue': Target Lock-on: Forcing player to turn to face someone who has a lock-on is disconcerting and frustrating. There needs to be another mechanism for indicating to a player when they are locked up, e.g. the player gets a HUD warning graphic with an indicator showing direction of attacker. ** This is working as intended by design. Design will investigate the possibilities *'C Issue': Target Lock-on: unable to use Darth vs Darth as player will not lock-on. ** For the purposes of the milestone, Darth Vader is on one ‘side’. Therefore a character attack on another character of the same side will not cause damage. The Imperial Officer was designated as being on an opposing side for the purpose of this Milestone. This is why Darth Vader vs Imperial Officer causes player damage ** LEC – understood, so in essence, we saw what would happen with friendlies, yes? However, we also noted that you couldn’t execute special melee moves unless you were locked on to a target – is that a bug? ** FRD: The lock on will happen if you attack *'C Issue': Target Lock-on: unable to damage other player unless locked on. ** Noted *'B Issue': Simple attack: currently, when R2 is pressed the model will do the base attack and return to the ready stance. Constant button-mashing does the attack/ready/attack/ready stance. Instead, button mashing should daisy chain different attack animations with each one flowing into the next for a fluid attack sequence. NOTE: reference SW Episode III game for example. ** This will be addressed for the next build ** LEC – great, will we see different attack animations? If so, will there be different block animations as counters to the attack animations? We can add this as a stretch goal for MS7. ** FRD: For the next build it will simply chain two existing animations. At this stage it is to primarily obtain gameplay feedback and approval, and not only visual goals. *'B Issue': There needs to be different attack styles for different Jedi/Sith, e.g. Vader/Emperor/Old Obi should have a ‘samurai’ technique while Luke, Anakin, etc have more acrobatic technique. ** Please can LucasArts provide a list of preferred styles for the range of Jedi / Sith characters for reference ** LEC – Definitely, we will put together a compilation of clips and categorize them into specific fighting styles. Our goal is to deliver you that compilation by the end of next week (10 Nov). *'B Issue': There needs to be different movement animations for different character types. Vader can’t use the same crouch, roll, crabwalk animations as other models. ** As with the previous statement – a list of preferred animations would be very useful ** LEC – Yes, they will fall into the same category as the fighting style, i.e. Old Ben wouldn’t be doing the rolls and jumps as young Ben, etc. *'B Issue': Block: player should be able to block and move at the same time. ** This will be added by the January 07 milestone deliverable *'C Issue': Block: what is the design for damage leaking through block? Can an attacker overpower a block? ** Yes, damage will seep through blocking. Overpowering / breaking a blocking stance is intended and will be under review in respect of gameplay balancing *'C Issue': Block: this uses stamina…what about attacks and big combos? ** This is a topic we feel would be best discussed in a conference call or in the forthcoming visits ** LEC – sounds good. STATUS: Approved as first pass on Jedi Melee Jedi Force Power (Lighting and Choke) prototype only affecting characters Overall *Good first pass on Lightning and Choke Good *Stance when using Choke Questions/Areas to Improve *'C Issue': Force Chock: Jedi using this needs an effect around the hands to indicate use of force power. ** Noted. Any reference that LucasArts can offer would be appreciated **'C Issue': Force Choke: Can/how does the victim break this? ** Button mashing / stick waggling is the current plan *'C Issue': Force Choke: Jedi victims need a new animation, e.g. an animation which has the victim drop their saber (which goes out) as they hold both hands to their neck – when the choke is broken they automatically force pull the saber which ignites in their hand and they immediately take the ready stance. ** This is something that requires discussion. There are technical considerations involved such as what happens to the lightsaber when it is dropped and the ramifications of using a pre-set animation to depict it dropping or using physics. ** LEC – understood, we can discuss next week. *'C Issue': Force Lightning: extremely powerful and long-ranged. Is the intent to tweak this via playtest? ** Yes, that is our intention *'C Issue': Force Lightning: When blocked by Jedi, the effect needs to be absorbed into their light saber. ** Noted. *'C Issue': Force Lightning: What is the visual target for the lightning effect? ** We intend to take the example that LucasArts supplied in an earlier milestone feedback report ** LEC – understood STATUS: Approved AI is functioning on ground/ in space (ability to follow player across VBF) Overall *Acceptable first pass on AI traversing terrain and capital ships Good *They generally get to their destination Areas to Improve *'B Issue': Pathfinding: AI get stuck on edges of ramps and cannot move ** Screenshots of these instances would be helpful. We will look to provide on-screen co-ordinates for later builds to aid this type of reporting. Pathfinding is currently work in progress ** LEC – we’ll focus on this when we get the coordinate function *'B Issue': Pathfinding: AI get stuck on side of starfighter and can neither enter vehicle nor move around it ** We’ll investigate this. Pathfinding is currently work in progress *'B Issue': AI animation when traveling needs to be improved – rather than having them running with gun extended, they should have a more natural stance as they walk or run to their destination. Upon encountering an opponent, they can then equip their weapon. **This is an opportunity to add humor to the encounter by having AI do various audio or animations to simulate surprise, etc. ** Animations are currently placeholder animations. Non-placeholder animations will be present by the end of 2006. We will investigate the suggestions raised in the final bullet-point STATUS: Approved All classes & playable units fully defined including attributes, weapons and role Overall *Good first pass on classes and playable units Good *Class structure feels as if it’s getting balanced Areas to Improve *'C Issue': Republic Heroes – they need non-Jedi Heroes *'C Issue': Heroes need to be balanced out across the different eras, e.g. same number of Jedi and non-Jedi heroes for each era. *'C Issue': Unlockable units – other than the Droideka, what other playable units are intended as unlockables for the other eras? *'C Issue': Han Solo special ability – consider a hero ability for Han Solo which when charged up allows him to frighten AI into stampeding away from him. **Once triggered, any AI within a set radius will run directly away from Han Solo **If player chases AI they will continue to run away **Han Solo animation will be as from Star Wars movie with him yelling as he chases them **Any AI encountered after ability is triggered will behave normally and engage Han Solo, which breaks the effect. At this point, all AI being chased will turn and engage Han Solo. **If player shoots while using ability, it breaks the effect as above. *** We would prefer to discuss these points in person when we meet at the start of November *** LEC – sounds good. STATUS: First pass approved Flying AI taking off / and landing on terrain and capital ships Overall *Good first pass on AI taking off/landing on terrain and capital ships Good *Landing on capital ships Areas to Improve *'B Issue': Landing on terrain: in some cases, the AI is unable to land on broken terrain – we had one instance of an AI stuck in the air on the side of a hill next to the Republic Base. ** This is a terrain issue. We will investigate and provide more feedback. *'B Issue:' This is a major issue with landing in general (for player controlled ships). **Main issues are ***Landing over broken or built up terrain ***Landing takes too long with need for vehicle to slow down, in many cases the the player will constantly press the land button which causes them to jump in and out of the ship. *** This is a terrain issue. We will investigate and provide more feedback. **What is the design for the player experience when landing a ship? *** We intend of the landing speed to be faster and less obtrusive to the flow of gameplay *** LEC – so, players will *not* be able to jump out of a vehicle instantly? The challenge is in conveying to the player the reason for the delay. *** FRD: Our ultimate aim is that as a hero, you can jump out instantly when the vehicle is in landing conditions. STATUS: Approved AI Can fly from ground to capital ship and vice versa Overall *Good first pass on AI traversing VBF Good *They maneuver from ground to ship well Areas to Improve *'B Issue:' Pathfinding: AI flying to ship appear to follow the same pre-determined path when going from the ground to the capital ship. What is the plan for giving the AI more natural flight paths? ** Key pathfinding nodes help direct AI ships to their destination. These are not hand placed by setup but infact generated at run time. The pathfinding system will be refined for the March 07 (FPP) build, where these will be much more realistically random. ** LEC – given the number of AI flying in the sky – what is the plan to manage their flight pattern with these pathfinding nodes. To set expectations: the game should be interesting to watch if I just let the AI run without any user interaction. Will we be able to make that assessment prior to the March milestone? ** FRD: As stated, the randomness will be worked on to ensure it is realistically random and therefore entertaining. It is unlikely that it be available for assessment before the March milestone. STATUS: Approved Destruction of vehicles with debris with placeholder effects Overall *Vehicles explode Good *Vehicles exploded with debris Areas to Improve *'C Issue:' Explosion effect – is the current graphical representation of the explosion a placeholder graphic? ** This is currently placeholder art *'C Issue:' Explosion graphical effect needs to be punched up with distortion rings or other effects which punctuate the violence of the explosions. ** These suggestions will be taken into account as the explosion art is refined *'C Issue': Different vehicles need different explosion effects. ** This will happen automatically to a degree. Gravity, speed, trajectory and mass will affect the appearance of how vehicles destruct. We do intend different classes of vehicles to have different effects, but this will not happen fully until later in the project ** LEC – actually, the reference is to different types of explosions as in tanks will explode with a red fireball type effect while fighters would have more of an anti-matter type explosion, etc. ** FRD: Agreed. As we said, we do intend different classes of vehicles to have different effects, but this will not happen fully until later in the project *'C Issue:' Debris – there is too much debris from some vehicle explosions, e.g. droid fighter. ** Lessening the debris is not a problem. Do LucasArts have a list of specific vehicles that suffer from too much debris? It should be possible to treat this on a vehicle by vehicle basis ** LEC – can vehicle destruction be added to the stage for each vehicle? That way we can view the effect and comment on it. ** FRD: Good idea. We will investigate it. *'C Issue:' Vehicle debris does not show up on the client machine. We understand that collision of the debris will not be networked, but we’d expect that both systems would see the debris from an explosion. ** This is a bug STATUS: Approved FX – re-entry in-game Overall *Nice first pass on saber effect Good *Looks very nice on the PS3 Areas to Improve *'C Issue:' Saber: trail effect – what’s the visual goal for this effect? ** The effect in the movies is the effect we are going for ** LEC – okay *'C Issue': The trail effect needs to fade towards the trailing edge instead of remaining a uniform opacity. ** We have based the trail effect as we saw it from movie footage. Which should we use? Movies or your comments? ** LEC – will review to ensure we’re on the same page. STATUS: Approved Comments/Bugs Bugs *'C Issue': Vehicles don’t appear to respawn in the capital ship’s hangar. ** Noted *'C Issue:' We were unable to fire our weapons while a vehicle is boosting. ** This is correct as per the current design. We will review this ** LEC – why shouldn’t players be able to fire weapons while boosting? ** FRD: The initial thoughts and design was that this offers a trade-off so that players will not boost all of the time their stamina bar allows. However, we are happy to consider and discuss alternatives. *'B Issue': Network **Occasionally when respawning in a network game, the game screen stays on the floating city view after the UI fades out. When this happens there’s no way to actually respawn on the ground and we must restart the game to continue. *** This is a known issue. Respawn code will be reviewed **Network play on the PC version often seems to get de-synched, and neither game clients, nor the server, seem to know that the game is not synched. See the example screenshot below, where one player sees a tank flying overhead. *** This should be corrected by the December build bug01.png *Vehicle debris does not show up on the client machine. We understand that collision of the debris will not be networked, but we’d expect that both systems would see the debris from an explosion ** This is a bug and will be fixed *'B Issue:' Graphics **On some machines, textures would all become black after playing for a minute or two (see screenshot). *** Did you mean PCs? What are the specs of the machines that suffer from this issue? Are they the machines detailed in the email entitled Battlefront III PC Multiplayer Testing Report? bug02.png Comments We would prefer to discuss these comment points in person during the visit *'C Issue:' Is the current spawn and class selection UI indicative of design/art direction? *'C Issue:' We need to remove the rocket head from the jetpack on the engineer. That’s an item that is more closely associated with Boba and Jango Fett and more probably belongs only on those models. *'B Issue:' Tatooine Level Design **The tank inside the Republic Base is difficult to get out of the base and into action ***All vehicles should have a direct path out of the ‘bases’. **Outpost CP? ***The area south of the Krayt Cavern – is that to be a CP? **Urban area ***The built up parts of the city are extremely tight and difficult to navigate with a tank – is it the intention to make this an infantry only area? ***SP objective mentions snipers on rooftops – does that mean some buildings will have interior spaces with staircases leading upwards? Was this objective designed when snipers had the jetpack? **Homestead CP ***In multiplayer, the CP is too open to camping especially from the rim above the CP. Moving the CP capture point to one of the anterooms might help alleviate this issue. ***In the wiki design – it shows the dune sea between the homestead and the city. However, in-game doesn’t have much distance. consider switching the Homestead location with the Sarlacc pit? **Mos Espa CP ***What is the intent for the landing pit? It looks larger than the pit in the movie. Will it be populated with objects to make infantry combat interesting? ***The spiralling ramp leading out of the pit is very steep – when running down the ramp the player character jumps instead of staying on the ground (issue in homestead ramp, too). ***Is it the intent to leave the CP in the middle of the landing pit? Can we put it into a more understandable setting such as near machinery or a computer console of some sort? ***What is the intent with both ramps into the pit area? Right now they are bland and featureless. Can we add some cover to the main ramp so that it’s not a big wide turkey shoot area? With the understanding that we want to avoid the omnipresent boxes and barrels of most shooters. **Cantina – is this going to be a CP on this level? ***Possible to have a bartender AI just minding the shop? Perhaps make it a health area where people can heal themselves...at the bar, of course. **What other reasons will the player have for going into the city? What is the extent of the playable area for the city, e.g. one edge is the landing pit one edge is the cantina? **Sarlacc Pit: Is the Sarlacc important for gameplay reasons or is it just a nice visual connection to the movies? Would be nice to make it have gameplay purpose (anyone falling into it dies) and then move it to an area which would get more usage. **CIS Landing Ship – that’s a huge ship...will it always land in the same place? ***There’s mention of this ship landing at the first captured command post...but that ship isn’t big enough to land near most of the CPs on Tatooine. *'B Issue:' CIS Interior **Given that the capital ship interiors will see quite a bit of play, we need to make them more interesting both from a gameplay perspective and from an aesthetic perspective. ***Right now it’s five rooms with large corridors connecting them. ***What are the plans for making the main rooms interesting for gameplay? ***What are the plans for making the corridors more interesting for gameplay, e.g. sections which jut out into hallway to provide cover? ***'NOTE': the one issue with level design is this: if a part of a level has been seen in a movie, then it must recreate that scene accurately. However, the flip-side of that is: we can be fairly creative with areas which have not been in the movies. ***We need railings along most catwalks to avoid the falling off syndrome of gameplay. Technical Comments General *Good to see the game up and running on the PS3! For the first PS3 delivery, the build seems quite stable. *We need to get better parity between the PC and X360 / PS3 builds, particularly with the renderer *Streaming (of textures) is extremely slow – taking several minutes before the high-res textures are loaded **Why has this slowed down significantly since the first build with streaming? ** Which platform is this comment referring to? ** FRD: Please can I have an answer to this. *If there are specific instructions on how to demonstrate a milestone deliverable, ie networking on PS3, please include detailed instructions with the milestone Performance *Performance is barely passable on all three platforms **PS3 performance is typically 4-7 FPS **With the exception of the arena on PS3, which is running at almost 60 FPS except when the players saber-lock *PS3 FPS display never drops below 5.00 FPS, but by firing rockets, I can get it down to approximately 2 FPS **Is the FPS display inaccurate, or are you limiting it so it never drops below 5 FPS? ** We believe there is a bug in the timings for the FPS counter at very low values. This will be investigated. *What are the major issues right now that are contributing to the performance problems? **Are you running everything on the PPU or X360 single core? ** No, we’re not running on single core. However, there is more we will be doing to improve performance in this area **Why is the PC build so much slower this milestone? ** There are considerably more buildings. More props are present. No visibility culling **It seems that you are CPU-bound, since when looking only at the sky dome, I still get 5-6 FPS on PS3 ** Noted *Are you doing any visibility culling at this stage? ** No. There is frustum culling on the props *I know it is early, and performance isn’t your top priority right now, but I want to get an understanding of what the issues are Renderer *There are major rendering artifacts on the X360 and PS3 **Z-buffer artifacts on PS3 (which you called out) **Incorrect or missing lighting **Lack of representative shaders, dropped geometry, etc. *When we visited FRD several months ago for due diligence, you demonstrated a fairly robust and mature PS3 and X360 renderer, with excellent lighting, shaders, full-screen FX and so on **Why are we seeing such a delta? ** BF3 uses the renderer for quite different purposes. There are numerous aspects unique to BF3 such as megatextures, transformation portals, area and object scale and vertical battlefront. ** LEC – do the BF3 renderer needs pose a problem to hitting the high visual target as established by other next gen titles? What steps will be taken to mitigate that risk? ** FRD: Please refer to the TDD. We are tackling some significant technical challenges and we have been targeting them since day one to get them all identified and dealt with. The risk is being minimized and managed. The Milestone Schedule and all deliverables were built with the recognized importance in investigating feasibility and maturing these features as early in the project as possible. **In some ways it appears that you are starting from scratch for BF3 based on the renderer capabilities. *Why are clouds, shadows, etc disabled on the console platforms? ** The clouds and shadows were absent from the PS3 build as they would have had an adverse affect on the build quality of this version Milestone 7 – Due November 20ᵗʰ, 2006 Overall Rating: ''PASS''